PLoS One. 2021 Mar 18;16(3):e0248338. doi: 10.1371/journal.pone.0248338. eCollection 2021.
PURPOSE: Regional-level measures can complement national antimicrobial stewardship programs. In Japan, sub-prefectural regions called secondary medical areas (SMAs) provide general inpatient care within their borders, and regional antimicrobial stewardship measures are frequently implemented at this level. There is therefore a need to conduct antimicrobial use (AMU) surveillance at this level to ascertain antimicrobial consumption. However, AMU estimates are generally standardized to residence-based nighttime populations, which do not account for population mobility across regional borders. We examined the impact of population in/outflow on SMA-level AMU estimates by comparing the differences between standardization using daytime and nighttime populations.
METHODS: We obtained AMU information from the National Database of Health Insurance Claims and Specific Health Checkups of Japan. AMU was quantified at the prefectural and SMA levels using the number of defined daily doses (DDDs) divided by (a) 1,000 nighttime population per day or (b) 1,000 daytime population per day. We identified and characterized the discrepancies between the two types of estimates at the prefectural and SMA levels.
RESULTS: The national AMU was 17.21 DDDs per 1,000 population per day. The mean (95% confidence interval) prefectural-level DDDs per 1,000 nighttime and daytime population per day were 17.27 (14.10, 20.44) and 17.41 (14.30, 20.53), respectively. The mean (95% confidence interval) SMA-level DDDs per 1,000 nighttime and daytime population per day were 16.12 (9.84, 22.41) and 16.41 (10.57, 22.26), respectively. The nighttime population-standardized estimates were generally higher than the daytime population-standardized estimates in urban areas, but lower in the adjacent suburbs. Large differences were observed in the main metropolitan hubs in eastern and western Japan.
CONCLUSION: Regional-level AMU estimates, especially of smaller regions such as SMAs, are susceptible to the use of different populations for standardization. This finding indicates that AMU standardization based on population values is not suitable for AMU estimates in small regions.